The interdisciplinary ViFaChem II project aims at providing a chemical digital library infrastructure for creating personalized information spaces. The value added services and scientific Web 2.0 techniques actively support chemical scientists and researchers in retrieval tasks as well as in deriving new knowledge from the collected information in a highly personalized fashion. The complex requirements of a digital library for chemists are described and an overall architecture tackling these requirements is presented. Also preliminary results regarding chemical entity recognition and automatic dynamic generated document facets are presented and discussed.